Technical Abstract In any population, with the presence of more than 100 genetic determinants, most breast cancer cases are likely to carry one, yet community occurrence is dominated by non-genetic risk factors. Adult identical twin women are representative of the population of origin in both genetic and non-genetic factors. However, when one gets breast cancer, the lifetime risk to the co-twin is much higher than expected on the basis of risk to an ordinary sister. Even so, only about a quarter of such co-twins become affected, and those that are become diagnosed years later. Thus affected pairs, whether breast cancer discordant or concordant, differ in the level of penetrance, and since the genetic determinants carried by concordant cases are unlikely to differ from those in discordant cases, the difference between such affected pairs, given the same genetic determinants, also reflects a different level of penetrance. We propose to verify the identity of genetic determinants between discordant and concordant pairs (using the Illumina OncoArray 500K), and to document the existence and timing of exogenous determinants of higher penetrance. This will be within discordant pairs, between the members of concordant pairs differing in age at diagnosis, and between cases from discordant and concordant pairs with the same or similar genotype. We will use adult twins' documented ability to accurately recall childhood differences between themselves in behaviors, environmental exposures, and given currant emphasis on adolescence, the rate of development. We will also use the Illumina HM450K BeadArray to evaluate whether the reported global hypo-methylation of WBCs from cases is attributable to the high risk genotype, and whether it persists after successful treatment.